⚛️ phenomenology

Supercool subtleties of cosmological phase transitions

This paper argues that the nucleation temperature is an unreliable indicator for cosmological phase transition completion, especially under strong supercooling, and advocates for using the percolation temperature and model-independent bounds on bubble wall velocity to accurately predict transition progress and gravitational wave production.

Original authors: Peter Athron, Csaba Balázs, Lachlan Morris

Published 2026-02-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Peter Athron, Csaba Balázs, Lachlan Morris

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the early universe as a giant, super-hot pot of soup. As it cools down, it's supposed to change its state, like water turning into ice. In physics, this is called a phase transition.

Usually, this happens smoothly. But sometimes, the universe gets "stuck" in a hot, metastable state (like supercooled water that hasn't frozen yet). It waits until it gets really cold before suddenly snapping into the new state. This is called a supercooled phase transition.

When this snap happens, it doesn't happen everywhere at once. Instead, little bubbles of the "new" universe (the true vacuum) start popping into existence inside the "old" universe (the false vacuum). These bubbles expand, crash into each other, and eventually swallow the whole universe. This violent collision creates ripples in space-time called gravitational waves, which we hope to detect with future telescopes like LISA.

For decades, physicists have used a specific rule of thumb to figure out if this transition happens and when to measure the gravitational waves. They call it the "Nucleation Temperature."

The Old Rule (The "One Bubble" Rule):
The old rule said: "The transition starts when, on average, one bubble pops into existence inside a 'Hubble Volume' (a giant cosmic sphere representing the observable universe)."
If you don't have at least one bubble in that giant sphere, the old rule says the transition fails. If you do, it succeeds.

The New Discovery (The "Supercooling" Twist):
This paper argues that the "One Bubble" rule is broken, especially when the universe supercools a lot. The authors, Peter, Csaba, and Lachlan, show that reality is much more subtle and surprising.

Here are the two weird scenarios they found:

1. The "Lonely Giant" (Scenario 1: Bubbles exist, but no transition)

Imagine you have a giant, empty warehouse (the Hubble Volume). You drop one giant balloon (a bubble) into it.

The Old Rule says: "Great! We have one bubble in the warehouse. The transition is happening!"
The New Reality: If that balloon is growing too slowly, it might never fill the warehouse before the universe expands so much that the balloon becomes insignificant. Even though you had "one bubble," the transition failed to complete. The universe stayed stuck in the old state.
The Lesson: Just having a bubble isn't enough; it has to grow fast enough to take over.

2. The "Magic Trick" (Scenario 2: No bubbles, but the transition happens)

Imagine the same warehouse. You drop zero balloons in.

The Old Rule says: "No bubbles? The transition failed. Nothing happened."
The New Reality: If the universe is supercooled enough, the few bubbles that do manage to pop into existence (maybe just a tiny fraction of a bubble per warehouse) can grow super-fast. They expand so quickly that they merge into a giant, connected web that swallows the whole universe, even though you never had "one full bubble" in the warehouse at the start.
The Lesson: You can have a successful transition even if you never hit the "one bubble" count.

Why Does This Matter?

1. The "Start Time" is Wrong:
Physicists use the "Nucleation Temperature" (when the first bubble appears) to calculate the strength of the gravitational waves. But if the transition can happen without that first bubble appearing (Scenario 2), or if the first bubble appears but the transition fails anyway (Scenario 1), then using that temperature gives you the wrong answer. It's like trying to predict the end of a race by looking at the starting line, but the runners took a shortcut or got stuck in traffic.

2. The "Percolation" Temperature is the Real Hero:
Instead of counting bubbles, the authors suggest we should look at Percolation.

Analogy: Imagine a crowd of people in a room. "Nucleation" is just counting how many people entered the room. "Percolation" is asking: "Is there a connected path of people from one side of the room to the other?"
The paper argues that the moment the bubbles connect and form a path across the universe (Percolation) is the real moment the transition happens. This is the best time to measure the gravitational waves.

3. The "Wall Speed" Matters:
The paper provides a new way to predict if a transition will succeed based on how fast the "walls" of the bubbles are moving.

If the walls move too slowly, the bubbles die out before they can take over.
If they move fast enough, even a tiny number of bubbles can win.
They give a mathematical "speed limit" for these bubble walls to ensure the universe actually changes its state.

The Big Picture

This paper is like a mechanic telling us that the "Check Engine" light (the old "One Bubble" rule) isn't always reliable. Sometimes the car is fine even if the light is off, and sometimes the car is broken even if the light is on.

By fixing how we measure these cosmic events, we can:

Stop throwing away good theories: We might have discarded some cool new physics models just because they didn't fit the "One Bubble" rule, even though they actually work.
Predict Gravitational Waves better: We can tell telescopes exactly when and how loud the "crash" of the bubbles will be, making it easier to find them.

In short: The universe is more creative and sneaky than we thought. It doesn't always follow the simple rules we wrote down in our notebooks, especially when it gets really cold!

1. Problem Statement

The paper addresses critical flaws in the standard methodology used to analyze first-order cosmological phase transitions, particularly those involving strong supercooling (where the universe remains in a metastable false vacuum state well below the critical temperature, $T_c$ ).

Current literature relies heavily on two concepts that the authors argue are misleading or insufficient for strongly supercooled transitions:

Unit Nucleation: The condition where, on average, one bubble nucleates per Hubble volume ( $N(T_n)=1$ ). This is often used as a proxy for the "start" of a transition and a guarantee that the transition will complete.
Nucleation Temperature ( $T_n$ ): The temperature at which unit nucleation occurs. This is frequently used as the reference temperature for calculating gravitational wave (GW) signals and baryogenesis.

The Core Issue: The authors demonstrate that these concepts decouple from the actual physical progress of the transition during strong supercooling. Specifically:

A transition can complete without ever satisfying the unit nucleation condition.
A transition can satisfy unit nucleation but fail to complete (percolate).
Simple heuristic approximations for $T_n$ break down in strongly supercooled regimes.
Using $T_n$ as a reference for GWs can lead to incorrect predictions of signal amplitude and frequency because the transition dynamics (bubble collisions) are governed by the percolation temperature ( $T_p$ ), not $T_n$ .

2. Methodology

The authors employ a combination of analytic derivations and numerical simulations to challenge standard assumptions.

Analytic Treatment

False Vacuum Fraction ( $P_f$ ): They utilize the standard formalism for the fraction of the universe remaining in the false vacuum, $P_f(t) = \exp(-V_{ext})$ , where $V_{ext}$ is the extended volume of bubbles.
Decoupling Events: They mathematically separate the event of Unit Nucleation (average 1 bubble/Hubble volume) from Percolation (formation of an infinite cluster of connected bubbles, $P_f \approx 0.71$ ) and Completion ( $P_f \to 0$ ).
Model-Independent Bounds: By assuming a Gaussian-like nucleation rate centered at $T_\Gamma$ and decomposing the energy density into radiation and vacuum components, they derive analytic bounds on the bubble wall velocity ( $v_w$ ). These bounds determine whether a transition will percolate or complete given a specific number of nucleated bubbles ( $N(0)$ ) and the ratio of the nucleation temperature to the radiation-vacuum equality temperature ( $T_\Gamma/T_{eq}$ ).
Simultaneous Nucleation Model: They use a simplified model (instantaneous nucleation) to derive analytic correction factors for various completion criteria.

Numerical Treatment

Models: They test their analytic predictions using two models:
1. M1 (Toy Model): A high-temperature expansion of the Standard Model with a zero-temperature potential barrier.
2. M2 (Real Scalar Singlet Model): A realistic extension of the Standard Model (SM + real scalar singlet) known for producing strong first-order phase transitions.
Benchmarks: They generate benchmark points where the number of bubbles nucleated per Hubble volume is tuned to be near unity ( $N \approx 1$ ).
Simulation Tools: They use modified versions of PhaseTracer, CosmoTransitions, and their own in-development code TransitionSolver to track phase evolution, calculate bounce actions, and determine false vacuum fractions without relying on nucleation heuristics.
Reheating Considerations: They discuss the impact of reheating (specifically for deflagrations and hybrids) on bubble growth and nucleation rates, noting that their primary analysis assumes detonations (where reheating is internal to the bubble).

3. Key Contributions

A. Identification of Two Counter-Intuitive Scenarios

The authors prove the existence of two scenarios that contradict standard intuition:

Scenario 1: Unit Nucleation without Percolation.
- Condition: $N(0) \ge 1$ but $P_f(0) > 0.71$ .
- Mechanism: Bubbles nucleate, but if the bubble wall velocity ( $v_w$ ) is too low (or the transition is too strongly supercooled), the bubbles expand too slowly to collide and merge before the universe cools further. The transition fails to complete despite "starting."
- Bound: They derive a bound on $v_w$ (Eq. 3.24) below which percolation is impossible even if unit nucleation occurs.
Scenario 2: Percolation without Unit Nucleation.
- Condition: $N(0) < 1$ but $P_f(0) \le 0.71$ .
- Mechanism: Very few bubbles nucleate, but if they nucleate early enough (high $T_\Gamma$ ) and expand rapidly (high $v_w$ ), they can grow to sizes larger than the Hubble volume and coalesce to cover the universe.
- Significance: This proves that the "one bubble per Hubble volume" rule is not a necessary condition for a successful phase transition.

B. Critique of Nucleation Heuristics

The paper evaluates common heuristics used to estimate $T_n$ (e.g., $S_3/T \approx 140$ ).

Failure Mode: These heuristics rely on Taylor expansions of the action around the nucleation temperature. In strongly supercooled transitions, the nucleation temperature often lies below the minimum of the action curve ( $T_n < T_{min}$ ).
Result: Heuristics either fail to find a solution, predict an incorrect $T_n$ , or yield action values that are physically inconsistent. The authors argue that direct numerical integration of the nucleation rate is necessary for accuracy.

C. Completion Criteria and Bubble Wall Velocity

The authors propose a hierarchy of completion criteria beyond just $P_f \to 0$ :

CC1: $P_f$ becomes negligible.
CC2-CC5: Conditions requiring the physical volume of the false vacuum ( $V_{phys} = a^3 P_f$ ) to decrease.
Finding: They find that satisfying CC1 (negligible $P_f$ ) is necessary but not sufficient. A transition can have $P_f \to 0$ asymptotically while the physical volume of the false vacuum continues to increase due to cosmic expansion.
Correction Factors: They provide empirical correction factors ( $\alpha_{CCi}$ ) to the bubble wall velocity bounds required to satisfy these stricter criteria. For example, to ensure $V_{phys}$ is decreasing at percolation (CC5), $v_w$ must be $\sim 2.23$ times higher than the velocity required merely for CC1.

D. Advocacy for Percolation Temperature ( $T_p$ )

The authors argue that $T_p$ (where $P_f \approx 0.71$ ) is the physically correct reference temperature for gravitational wave production because:

It marks the onset of significant bubble collisions.
It is directly connected to the transition progress.
In strongly supercooled transitions, $T_n$ and $T_p$ can differ by tens of GeV, making $T_n$ a poor proxy for GW signal characteristics.

4. Results

Validation of Scenarios: Numerical benchmarks in both M1 and M2 confirmed that Scenario 1 (nucleation without percolation) and Scenario 2 (percolation without nucleation) are physically realizable.
Velocity Bounds: The analytic bounds on $v_w$ derived in Section 3 matched the numerical thresholds for percolation and completion with high precision across different regimes (radiation vs. vacuum domination).
Hubble Volume Exceedance: The authors proved that in strongly supercooled transitions, individual bubbles can grow larger than the Hubble volume ( $R_{bubble} > H^{-1}$ ). This invalidates the intuition that "one bubble per Hubble volume" is required for the transition to span the universe.
Heuristic Breakdown: The study showed that for strong supercooling ( $\delta_{sc} \gtrsim 0.5$ ), standard heuristics for $T_n$ deviate significantly from the true value or fail entirely.
Numerical Uncertainty: The paper highlights that numerical errors in bounce action solvers (like CosmoTransitions) can obscure the precise determination of $T_n$ in extreme supercooling, necessitating tighter tolerances and specialized handling of thin-wall limits.

5. Significance

Correcting GW Predictions: By shifting the reference temperature from $T_n$ to $T_p$ and accounting for the possibility of transitions completing without unit nucleation, the paper prevents the misclassification of viable models. This is crucial for interpreting future data from detectors like LISA.
Refining Model Constraints: Many extensions of the Standard Model (e.g., for baryogenesis or dark matter) rely on strong supercooling. Previous studies might have incorrectly discarded viable parameter points because they failed the "unit nucleation" test, or conversely, accepted points that would not actually complete the transition.
Methodological Rigor: The paper advocates for a more rigorous approach to phase transition analysis, moving away from simplified heuristics toward full numerical integration of the false vacuum fraction, especially for strongly supercooled scenarios.
Tool Development: The authors introduce TransitionSolver, a tool designed to efficiently calculate percolation and completion temperatures without the computational cost of brute-force integration, facilitating large-scale parameter scans.

In summary, this work fundamentally challenges the "unit nucleation" dogma in cosmology, demonstrating that the dynamics of strongly supercooled phase transitions are governed by bubble growth rates and percolation thresholds rather than the simple count of nucleated bubbles.